Abstract 3421: Delayed Diagnosis of Cerebellar Infarctions

Abstract

Background: An acute ischemic cerebellar stroke presents a clinical and radiological diagnostic challenge, which may be due in part to the difficulty in recognizing its symptoms and signs and to an over-reliance on CT. This retrospective study addresses for the first time the differences in clinical features and outcomes between those individuals with a cerebellar infarct who were correctly diagnosed (CD) on initial presentation compared to those who were initially misdiagnosed (MD), and consequently had a delay in correct diagnosis. Methods: A retrospective review was conducted of our stroke registry from 09/2003 to 07/2010. Forty three patients had an isolated cerebellar infarction confirmed by MRI. Common neurological signs and symptoms were chosen to analyze their association with correct diagnosis or misdiagnosis. Misdiagnosis was defined as the diagnosis given by the first physician who saw the patient in the outpatient or inpatient setting. Onset to door time is defined from onset of the patient’s symptoms to our ED. The primary outcome was mRS at discharge. Data was analyzed using the chi square test, Fischer exact test, Wilcoxon-rank sum test and logistic regression. Results: Sixteen (37%) patients were MD; among these patients, the median duration of delay to being CD was 24 hours (IQR 9-48h). The median onset to our ED was 11.1 hrs in the MD versus 5.5 hrs in the CD group (p=0.02). The median age in the MD group was 49 y/o versus 65 y/o in the CD group (p< 0.001). More patients with hypertension were CD (71.4%, p<0.05). Weakness was associated with a decreased likelihood of MD (OR 0.18; 95% CI 0.034-0.94). The presence of nausea increased the tendency of MD (OR 5.6; 95% CI 1.06-29.6), and a trend toward MD was present if a patient had vertigo (OR 3.3; 95% CI 0.92-12.1). Five (18.5%) patients who were CD received tPA compared to none in the MD group (p=0.14). Ten (37%) MD patients presented within the time window for tpa, and subsequently were no longer candidates for thrombolysis. There was a trend toward a good outcome at discharge (mRS 0-1) in the CD group (OR 3.5; 95% CI 0.9-13.8). There were no deaths before discharge in either group. Conclusion: Timely identification of an acute cerebellar infarction requires the recognition that stroke may occur in younger individuals and that symptoms of cerebellar stroke are commonly associated with more benign processes such as gastroenteritis. Although limited by sample size, our data suggests that timely diagnosis of cerebellar infarction might increase thrombolysis rates and may impact functional outcome.